Abstract:
This paper presents an experimental study of a Probabilistic
Road Map (PRM) based obstacle avoiding algorithm, for
motion planning of a non-holonomic mobile robot in cluttered
dynamic environment. The PRM approach uses a fast and
simple local planner to build a network representation of the
configuration space. It is trading off the distance to both
static objects and moving obstacles in compute the travelled
path. Our work has been implemented and tested on Player /
Stage, real time robotic software, in extensive simulation
runs. The different experiments that runs had demonstrate
that our approach is well suited to control the motions of a
robot in a cluttered environment and demonstrates its
advantages over other techniques.
